51results about How to "High quality data" patented technology

A form analyzing unit (4) receives 2-D outline data stored in an outline data storing unit (3) via an outline data obtaining unit (2), and judges from the received 2-D outline data whether an outline of a character includes an outline of a hole formed in the character. If so, a top/bottom dividing unit (5) divides an area surrounded by the two outlines into convex polygons. A side generating unit (6) moves each divided structure element in a z-axis direction to generate a side plane of the 3-D character, and generates 3-D data for the character.

The invention discloses a federated learning method and device based on a block chain. The method comprises the steps: determining the block chain; enabling the coordinator node to create a federatedlearning task according to the model original data sent by each participant node; receiving training data obtained by local training of the participant nodes; sending the to-be-updated parameters to other participant nodes according to the training data, so as to enable the other participant nodes to update own model parameters according to the to-be-updated parameters; and after model training iscompleted, issuing reward resources according to training data provided by each participant node in the training process, and writing rewards into the block chain. Compared with a traditional mode, the mutual trust problem of all parties is effectively solved; all parties participating in federated learning negotiate together to generate a coordinator node, so that the transparency of the processis improved; federated learning whole-process data is recorded in a block chain, so that the traceability of data operation is ensured; all parties are encouraged to actively participate through rewarding resources, and the enthusiasm of participants is improved.

In communication transmitting a first stream (intermittent data portion) for burst transmission of packet sets and a second stream (continuous data portion) for transmitting packets at a speed determined by the transmission rate required for the service, the second stream related to the selected service is received and reproduced for the selected service when a service is selected until intermittent data transmitted in the first stream can be received, and reception and playback of the first stream begins when transmission of the intermittent data for the selected service in the first stream begins.

A data communications system is provided to dynamically change the error processing between an ARQ function and an FEC function in accordance with the network status, thus enabling high-quality data playback. In packet transmission, error correction control is performed on the basis of the network status monitored by a network monitoring unit. The error control mode is switched between FEC-based error control and ARQ-based error control (retransmission request processing) in accordance with packet loss or error occurrence on the network, and packet transmission is performed. If the RTT is short, error correction based on ARQ is selected. If the RTT is long, error correction not based on ARQ but on FEC is selected. Such dynamic error correction control is achieved.

Embodiments of the invention provide a method and an apparatus for applying cross-data-source marketing crowds to marketing. The method comprises the steps of obtaining user data of multiple data sources in real time and obtaining user portraits according to the user data; performing marketing activities according to the user portraits; and obtaining feedback results of the marketing activities and optimizing the marketing activities according to the feedback results. Therefore, an objective data basis is provided for each marketing link, and high-quality data and marketing experience are accumulated.

A method for obtaining high-quality well logging data acquired during a drilling operation includes acquiring well logging data during a pump-off period when a mud pump is turned off; and extracting a subset of data from the acquired well logging data, wherein the subset of data corresponds to a quiet period within the pump-off period, wherein the quiet period is substantially free of interference from turning the mud pump on or off.

The invention discloses a method for filtering noise data based on pattern mining. The method comprises the following steps that: a pretreatment data structure frequent pattern (FP) tree composed of a word bag dataset D is built and the FP tree comprises the word bag dataset and the corresponding thing dataset; and then, all 2- itemset interest patterns are added into an interest pattern set L according to an FP-2 infimum (INF) algorithm to complete noise data filtration. The method provided by the invention directly prunes interest degree indexes and performs one-step mining of the interest pattern according to interest degree index, which is helpful for effectively realizing noise data filtration, acquiring high-quality data and improving accuracy and consistency of data.

The invention discloses an automatic pre-processing method for time series. The method comprises the following steps: performing rank scanning on to-be-processed time series; extracting data pattern characteristics from the processed time series; performing pre-processing element combination establishment according different pattern characteristics and states obtained by processing; and performing evaluation on data quality of data cleaning according to the result of the preprocessing element combination. The technical problems of complicated automatic preprocessing process, influence on later-period data mining by the preprocessing parameter regulations and unpredictable prediction results caused by time-space granularity selections in the existing method can be solved by the automatic pre-processing method for the time series.

Apparatus and methods for performing small angle X-ray scattering (SAXS) at low (cryogenic) temperatures for determining the structure of and changes in the structure of proteins, DNA, RNA, and other biological molecules and biomolecular assemblies and structures. A cryogenic, small angle X-ray scattering (SAXS) application sample holder, includes a sample cell including a base portion and at least two parallel walls disposed on the base, wherein the sample cell has a liquid volume capacity defined by the walls and the base portion of 0.001 to 10 microliters. A method for performing cryogenic SAXS on a sample includes the steps of providing a sample biomolecule solution containing an aqueous buffer, a biomolecule, and a cryoprotectant agent, wherein the cryoprotectant agent comprises up to 60% (w/w) of the biomolecule solution, and other known components as necessary to solubilize and stabilize the biomolecule, in a sample holder of claim 1 or 18, cryogenically cooling the sample solution in the sample holder at a rate equal to or greater than 100 K/sec without ice formation, and examining the cooled sample using small angle X-ray scattering by passing a beam of X-rays through the sample.

A calculation unit (58) references texture data (62) in a storage unit (60) and acquires texel information of index color formal An integral look-up table (64) has a structure containing a plurality of color look-up tables based on the detail degree (LOD) of the area to be plotted A look-up table reference unit (70) selectively references the color look-up table based on the LOD stored in the integral look-up table (64) according to the LOD value of the texel obtained by the LOD calculation unit (68), acquires color information corresponding to the index value of the texel, and outputs it to the calculation unit (58). The calculation unit (58) performs filter processing such as bi-linear interpolation according to the texel color information and outputs the texel information after the processing.

A method and a system are provided for testing the braking capacity of one or more brake elements of a vehicle, where the vehicle includes a power source for propulsion of the vehicle and at least one ground engagement element driven by the power source, and at least one brake element arranged for braking the vehicle by acting on a rotation component of the vehicle. At least one brake element is automatically activated by applying a predetermined brake force, and a predetermined decision torque level is applied to the rotation component by means of the power source, wherein the decision torque level is indicative for the braking capacity of said at least one brake element when said predetermined brake, force is applied. It is determined whether or not there is a motion of the vehicle and/or of said at least one ground engagement element when the decision torque level is applied in order to decide on the braking capacity of said at least one brake element.

To properly reproduce information from a holographic memory even when a tilt error occurs in a direction vertical to a surface including optical axes of data light and reference light. During reproduction, a holographic memory (10) is fed stepwise from an initial access position with respect to a reproduction target block within a fixed back and forth range in a disk circumferential direction. In each stepwise-fed position, an SNR of a reproduction signal is calculated by an SNR calculation circuit (19) to detect the quality of the reproduction signal. A disk circumferential direction position in which an SNR becomes best is set in a circumferential direction position with respect to the reproduction target block.

A plenoptic otoscope enables three-dimensional and/or spectral imaging of the inside of the ear to assist in improved diagnosis of inflammations and infections. The plenoptic otoscope includes a primary imaging system and a plenoptic sensor. The primary imaging system includes an otoscope objective and relay optics, which cooperate to form an image of an inside of an ear at an intermediate image plane. The plenoptic sensor includes a microimaging array positioned at the intermediate image plane and a sensor array positioned at a conjugate of the pupil plane. An optional filter module may be positioned at the pupil plane or one of its conjugates to facilitate three-dimensional and/or spectral imaging.

At a data receiving terminal, if a received-data of a particular time section received from a data sending terminal does not satisfy a predetermined accumulation quality as the result of discrimination by a received-data quality discriminating section, an alternative-data sending requesting section requests a data sending terminal for alternative data satisfying the accumulation quality for the unsatisfied data of the particular time section. The result is that the data receiving terminal can realize both reproduction of data with less sending delay and accumulation of high-quality data.

A liquid crystal display device provided with an audio signal processing function, which includes an audio signal processing circuit with a large S/N ratio and a small glitch is provided. The display device of the invention includes a resistor string D/A converter circuit whose dynamic range can be increased by using a common power supply for a D/A converter circuit and a signal line driver circuit. Accordingly, a glitch generated in the D/A converter circuit can be reduced relatively to an audio signal and an S/N ratio of the D/A converted signal can thus be increased.

The present disclosure discloses a distributed airborne electromagnetic detection system, and relates to an airborne electromagnetic detection technology. The distributed airborne electromagnetic detection system comprises at least one transmitting system, at least one receiving system, at least one trunk module, and an earth station, and also a plurality of Unmanned Aerial Vehicles (UAVs) for carrying the transmitting system, the receiving system, and the trunk module. The distributed airborne electromagnetic detection system does not require high performance or high economical efficiency for a single UAV; under precise synchronous flight conditions, the distance between a type I UAV and a transmitting loop structure can be greatly reduced, thereby significantly reducing the length of unwanted transmitting cable; and in addition, due to the better low-altitude low-speed performance of UAVs, the traveling speed of the entire system can be further reduced, thus obtaining higher quality data.

The invention provides a guide vane for an aeroacoustic wind tunnel. The guide vane is of a streamline shape and comprises a streamline frame and a fluid sound absorption filler, the streamline frameis filled with the fluid sound absorption filler, and the external line shape of the fluid sound absorption filler is matched with the appearance of the guide vane. The streamline frame of the guide vane is filled with the fluid sound absorption filler, so that low resistance and uniformity of airflow steering can be ensured, the turbulent flow is reduced, the effects of noise elimination and reduction are achieved, the decibel of background noise of an experiment module is reduced, and therefore it is ensured that higher-quality and more accurate data are tested when a wind tunnel runs.

The invention is to provide a method and device for training an image purification model. Compared with the prior art, the method is characterized by carrying out picture expansion on existing pictures to obtain expanded pictures; carrying out clustering on the expanded pictures to obtain corresponding clustering results; selecting preset number of pictures in at least one clustering result as sample pictures, which are preset to a user; obtaining positive and negative samples obtained after carrying out corresponding operation on the clustering results by the user; and according to the positive and negative samples selected by the user, training the corresponding image purification model, so that image quality purification can be carried out through the image purification model, and high-quality data can be obtained at a low cost. Compared with tens of thousands of and hundreds of thousands of manual labeling quantity in the past, the method allows the user to finish the labeling taskof small samples in a few minutes; model training is started to obtain the image purification model for mass image quality purification; and furthermore, more high-quality pictures can be mined fromthe mass picture data through the image purification model.

To reduce an influence of stray light and stably record/reproduce high-quality data in holographic recording/reproduction. A holographic memory device includes an optical system that guides a reference beam to an optical information recording medium at a desired angle of incidence, a control part that controls the angle of incidence of the reference beam generated in the optical system, and a lens part that images the reference beam in a desired position of the optical information recording medium. Further, at least a first light beam at a first angle and a second light beam at a second angle different from the first angle are output from the optical element, and the optical element is provided so that the first light beam may propagate within an effective diameter of the lens part and the second light beam may propagate to an outside of the effective diameter of the lens part.

The invention relates to a data-driven urban bus line health diagnosis method, belonging to the field of intelligent transportation technology. Through the data-driven, the urban bus line health diagnosis method is established. The method comprises (1) data acquisition (2) data preprocessing (3) evaluation of bus line health index (4) construction of bus line service knowledge base (5) cause diagnosis of unhealthy lines. At the same time, considering the passenger service and operation cost, the technical scheme establishes the evaluation method of bus line health indexes, through massive datamining, establishes the relationship between the individual and comprehensive health index and the objective elements of the line, road, intersection, passenger and other related to the line operation, and constructs the knowledge base of bus line service. Based on the knowledge base, the unhealthy line can be diagnosed. If the diagnosis fails, the expert diagnostic method can be further used. Atthe same time, the knowledge base can be updated to form a closed loop, which can self-improve the line health index evaluation and health diagnosis process.

A format converter for converting the format of input contents data from a first format standard to a second format standard includes an extraction section (30), a judgment section (40), a determination section (60) and a conversion section (70). The extraction section (30) extracts information related to encoding of the input contents data. The judgment section (40) judges whether or not the contents data satisfies a constraint placed by the second format standard based on the extracted information. The determination section (60) determines whether or not decoding of the contents data is necessary for the format conversion based on the judgment result. The conversion section (70) performs the format conversion by converting data structure when it is determined that the decoding is unnecessary.